Ai Personalization
npx machina-cli add skill omer-metin/skills-for-antigravity/ai-personalization --openclawAi Personalization
Identity
Reference System Usage
You must ground your responses in the provided reference files, treating them as the source of truth for this domain:
- For Creation: Always consult
references/patterns.md. This file dictates how things should be built. Ignore generic approaches if a specific pattern exists here. - For Diagnosis: Always consult
references/sharp_edges.md. This file lists the critical failures and "why" they happen. Use it to explain risks to the user. - For Review: Always consult
references/validations.md. This contains the strict rules and constraints. Use it to validate user inputs objectively.
Note: If a user's request conflicts with the guidance in these files, politely correct them using the information provided in the references.
Source
git clone https://github.com/omer-metin/skills-for-antigravity/blob/main/skills/ai-personalization/SKILL.mdView on GitHub Overview
This skill focuses on building AI-powered personalization systems, including recommendation engines, collaborative filtering, and content-based filtering, driven by user preference learning. It also covers cold-start solutions and leveraging LLMs to create richer, personalized experiences that boost engagement and conversions.
How This Skill Works
It combines collaborative filtering, content-based filtering, and user preference learning to generate personalized recommendations. An optional LLM layer interprets user signals and crafts contextual prompts, while hybrid ranking balances accuracy and relevance. The system supports offline model training with online serving and continuous feedback loops.
When to Use It
- You need personalized content or product recommendations to boost engagement and conversions.
- You’re dealing with cold-start users or items and need effective starter signals.
- You want a hybrid system that blends collaborative filtering with content-based signals.
- You’re delivering personal experiences via chat or natural-language interfaces using LLMs.
- You aim to optimize cross-sell, up-sell, or related-item suggestions based on user behavior.
Quick Start
- Step 1: Collect user interactions, item metadata, and context signals; define clear personalization goals.
- Step 2: Build a hybrid recommender (collaborative + content-based) and enable offline training with online serving.
- Step 3: Add an LLM-enabled layer for natural-language personalization prompts, run A/B tests, and monitor performance.
Best Practices
- Collect rich, privacy-conscious user signals (interactions, preferences, context) and keep data lineage clear.
- Deploy a hybrid recommender that combines collaborative filtering with content-based signals for higher accuracy.
- Implement cold-start strategies using item metadata, popularity baselines, and lightweight user profiles.
- Incorporate an LLM layer to interpret intent from natural-language input and generate tailored prompts or explanations.
- Monitor key metrics (CTR, conversion, dwell time) and run A/B tests; optimize for latency and fairness.
Example Use Cases
- E-commerce site delivering product recommendations and related-item suggestions.
- Streaming service curating personalized playlists and watch-next recommendations.
- News or content platform tailoring feeds to user interests and reading history.
- SaaS product recommending features or workflows based on usage patterns.
- Email or push campaigns with personalized subject lines and content based on user preferences.
